Environmental Engineering Reference
In-Depth Information
hair mercury concentrations were 2.55 ppm in men and
1.43 ppm in women; 92% of men and 72% of women had
levels greater than 1.0 ppm (Yasutake et al., 2003). Average
hair mercury in Japanese and Korean communities in the
state of Washington measured six and three times, respec-
tively, the national level (geometric means, 1.23, 0.61, and
0.2 ppm, respectively), and fi sh consumption was similarly
elevated relative to national levels (Tsuchiya et al., 2008).
Fish is an important component of the diet in some Native
American cultures as well, and average hair mercury con-
centrations over 5 ppm have been measured among Cree
Indians living in Northern Quebec (McKeown-Eyssen and
Ruedy, 1983). The relatively high mercury levels in the fi sh-
eating populations living in the Faroe Islands, the Republic
of the Seychelles, and the Amazon region of Brazil were
described previously (Grandjean et al., 1992; Myers et al.,
2003), and many more examples exist, particularly from
island populations (Mahaffey et al., 2009).
Thus far, Korea is the only Asian country that has
begun ongoing population-based biomonitoring for mer-
cury exposure (Son et al., 2009). Preliminary results for
2007-2008—available for a subset of 2342 adult study
participants—are consistent with other studies that have
measured blood mercury concentration in Asians. The
geometric mean was 3.8 µg/L and the 95th percentile was
14.94 µg/L. There was also a pattern of increasing mercury
exposure with increasing socioeconomic status that was
attributed to greater fi sh consumption.
In the United States, people with higher education and
socioeconomic status may also have higher blood and hair
mercury levels due to consumption of more expensive,
predatory fi sh (e.g., swordfi sh and sushi-grade tuna), which
tend to be higher in mercury (Hightower and Moore, 2003;
Knobeloch et al., 2005; McKelvey et al., 2007). Approximately
one in six patients visiting a middle- to high-income inter-
nal medicine clinic in San Francisco was deemed likely to
be ingesting methylmercury above the EPA's RfD. Indeed,
among those selected for their high fi sh consumption, 89%
had total blood mercury concentrations greater than or equal
to 5.0 µg/L; 16% had levels of at least 20 µg/L (Hightower
and Moore, 2003). Swordfi sh consumption was most strongly
correlated with elevated mercury levels. The seven children
included in this study had blood mercury levels from 11 to
26 µg/L and hair levels ranging from 3 to 15 ppm.
associated with higher urine mercury levels (Apostoli et al.,
2002; Levy et al., 2004; McKelvey et al., 2010).
The GerES fi rst collected population-based biomonitoring
data on mercury in urine as part of its 1985 survey. In 1998,
the geometric mean urine mercury concentration, measured
in 4730 adults 18-69 years old from both parts of the country
was 0.34 µg/g creatinine (95th percentile of 2.0 µg/g) (Becker
et al., 2003). The geometric mean was highest (0.89 µg/g)
among those who had mercury amalgams on more than
eight teeth. The geometric mean urine mercury concentra-
tion measured in 1354 children 6-14 years of age was only
0.10 µg/L (95th percentile of 0.52 µg/L) in 2003-2006, which
was statistically signifi cantly lower than the 1990-1992 esti-
mate of 0.54 µg/L (95th percentile of 3.9 µg/L) (
p
value for
the difference between geometric means, <0.01), most likely
because Germany eliminated the use of mercury amalgams
for dental restoration in children during this period (Schulz et
al., 2007b). Levels measured in 619 Czech children in 2001-
2003 were similar to earlier German levels (geometric mean,
0.45 µg/g; 95th percentile, 4.18 µg/g) (Batariova et al., 2006).
The most recent data on urine mercury concentrations
in the US population come from, 2001-2002 NHANES
data. The geometric mean was 0.62 µg/g creatinine (95th
percentile, 3.0 µg/g) based on data from 1960 women 16
to 49 years of age (CDC, 2005). Estimates were higher
than in the German population, but lower than levels
measured among 160 adult female blood donors in the
Czech Republic in 2001-2003 (95th percentile, 11.8 µg/g)
(Batariova et al., 2006). In the Czech study, women had
signifi cantly higher urine mercury levels than men. The
NYC HANES identifi ed elevated urine mercury levels in
a subgroup of Dominican women, some of whom were
later found to have been using mercury-containing skin-
lightening creams (Mckelvey, 2011).
Future Biomonitoring Efforts
Biomonitoring for a chemical is most appropriate when there
are suspected health effects and uncertain exposure patterns
in the population, when an adequately accurate and eco-
nomically feasible laboratory test exists, and when there are
known interventions for reducing population risks. Mercury
meets these criteria and therefore is considered a high prior-
ity on many governmental agendas. The European Union
has included biomonitoring for mercury in its Environment
and Health Action Plan 2004-2020 (Smolders et al., 2008),
and Canada has included it in its Canadian Health Measures
Survey (Wong and Lye, 2008). The US NHANES continues
to conduct mercury biomonitoring in blood and urine from
children and adults, both male and female (CDC, 2012). At a
state level, California and Minnesota have passed legislation
to develop biomonitoring programs, and mercury is a pri-
ority chemical (California Offi ce of Environmental Health
Hazard Assessment, 2009; Minnesota Department of Health,
2009). Korea is the only Asian country that has begun a
national biomonitoring effort.
Mercury Levels in Urine
Urine mercury concentration in the general population
is most consistently associated with the presence of mer-
cury amalgams used for dental restoration. Other exposure
sources, including mercury-containing skin-lightening
creams, traditional medical remedies, and ritualistic prac-
tices, may also be present in some subgroups of the popula-
tion. Perhaps because of demethylation of methylmercury
in vivo, or the concurrent presence of inorganic mercury
in fi sh tissue, frequent fi sh consumption has also been
